Source: US Environmental Protection Agency EJSCREEN Tool, 2020 data (last modified 7/1/21)
EJSCREEN is an “environmental justice (EJ) mapping and screening tool” produced by the EPA.
glimpse(ejscreen)
## Rows: 155
## Columns: 36
## $ ID <dbl> 510030101001, 510030101002, 510030101003, 510030102011, 510…
## $ PRE1960PCT <dbl> 0.071991001, 0.298299845, 0.256756757, 0.027385892, 0.08238…
## $ DSLPM <dbl> 0.1316935, 0.1316935, 0.1316935, 0.1974282, 0.1974282, 0.21…
## $ CANCER <dbl> 24.95765, 24.95765, 24.95765, 28.33822, 28.33822, 28.15892,…
## $ RESP <dbl> 0.3120049, 0.3120049, 0.3120049, 0.3656255, 0.3656255, 0.35…
## $ PTRAF <dbl> 0.00000000, 0.00000000, 0.31287689, 286.71433099, 4.0751412…
## $ PWDIS <dbl> 0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00, 0.0…
## $ PNPL <dbl> 0.03555328, 0.05853532, 0.06053500, 0.02780625, 0.03162721,…
## $ PRMP <dbl> 0.05435428, 0.09491017, 0.14614959, 0.05496790, 0.04548612,…
## $ PTSDF <dbl> 0.10236159, 0.06039390, 0.08200253, 0.21222923, 0.23416369,…
## $ OZONE <dbl> 41.64513, 41.64513, 41.64513, 41.65850, 41.65850, 41.65348,…
## $ PM25 <dbl> 7.241029, 7.241029, 7.241029, 7.386133, 7.386133, 7.364257,…
## $ P_LDPNT <dbl> 33.48404, 62.76819, 58.92123, 21.17666, 35.58471, 51.33492,…
## $ P_DSLPM <dbl> 7.637395, 7.637395, 7.637395, 18.744228, 18.744228, 21.5940…
## $ P_CANCR <dbl> 22.04262, 22.04262, 22.04262, 35.76293, 35.76293, 34.95291,…
## $ P_RESP <dbl> 19.50741, 19.50741, 19.50741, 32.45459, 32.45459, 30.91068,…
## $ P_PTRAF <dbl> 0.000000, 0.000000, 5.341820, 56.287080, 8.495607, 7.452488…
## $ P_PWDIS <dbl> 0.00000, 0.00000, 0.00000, 0.00000, 0.00000, 0.00000, 0.000…
## $ P_PNPL <dbl> 31.93691, 48.19756, 49.36919, 25.50034, 28.75482, 36.94343,…
## $ P_PRMP <dbl> 5.555356, 14.571876, 26.639997, 5.670226, 3.932652, 8.65649…
## $ P_PTSDF <dbl> 14.103431, 7.377108, 10.976983, 26.610963, 28.217063, 21.23…
## $ P_OZONE <dbl> 40.64001, 40.64001, 40.64001, 40.75917, 40.75917, 40.71748,…
## $ P_PM25 <dbl> 16.10605, 16.10605, 16.10605, 18.16651, 18.16651, 17.83501,…
## $ T_LDPNT <chr> "0.072 = fraction pre-1960 (33%ile)", "0.3 = fraction pre-1…
## $ T_DSLPM <chr> "0.132 ug/m3 (7%ile)", "0.132 ug/m3 (7%ile)", "0.132 ug/m3 …
## $ T_CANCR <chr> "25 lifetime risk per million (22%ile)", "25 lifetime risk …
## $ T_RESP <chr> "0.31 (19%ile)", "0.31 (19%ile)", "0.31 (19%ile)", "0.37…
## $ T_PTRAF <chr> NA, NA, "0.31 daily vehicles/meters distance (5%ile)", "290…
## $ T_PWDIS <chr> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, "0.0000…
## $ T_PNPL <chr> "0.036 sites/km distance (31%ile)", "0.059 sites/km distanc…
## $ T_PRMP <chr> "0.054 facilities/km distance (5%ile)", "0.095 facilities/k…
## $ T_PTSDF <chr> "0.1 facilities/km distance (14%ile)", "0.06 facilities/km …
## $ T_OZONE <chr> "41.6 ppb (40%ile)", "41.6 ppb (40%ile)", "41.6 ppb (40%ile…
## $ T_PM25 <chr> "7.24 ug/m3 (16%ile)", "7.24 ug/m3 (16%ile)", "7.24 ug/m3 (…
## $ AREALAND <dbl> 74984742, 227406673, 60429005, 39270835, 30830002, 28336114…
## $ AREAWATER <dbl> 463193, 501595, 852995, 367647, 395358, 844351, 1021277, 50…
Observations are block group estimates of key environmental indicators:
PRE1960PCT)DSLPM and PM25)CANCER)RESP)PTRAF)PNPL)PRMP)PTSDF)OZONE)PWDIS)P_ indicates percentile ranks for each variable, and T_ indicates map popup text.
meta %>%
mutate(label = paste0(varname, ": ", about)) %>%
select(label) %>%
as.list()
## $label
## [1] "ID: 12-digit FIPS block group code"
## [2] "PRE1960PCT: % of housing built before 1960 -- lead paint indicator"
## [3] "DSLPM: Diesel particulate matter level in the air, measured in micrograms per cubic meter"
## [4] "CANCER: Cancer risk due to toxics in the air"
## [5] "RESP: \"Ratio of exposure concentration to health-based reference concentration\""
## [6] "PTRAF: Average number of daily vehicles at major roads divided by distance in meters"
## [7] "PWDIS: Toxicity-weighted stream concentrations divided by distance in kilometers"
## [8] "PNPL: Number of National Priorities List (NPL) sites within 5 km divided by distance in kilometers"
## [9] "PRMP: Number of Risk Management Plan (RMP) facilities within 5 km divided by distance in kilometers"
## [10] "PTSDF: Number of Treatment Storage and Disposal (TSDF) facilities within 5 km divided by distance in kilometers"
## [11] "OZONE: Summer daily average of ozone concentration in the air, in parts per billion"
## [12] "PM25: Yearly average PM2.5 level in the air, measured in micrograms per cubic meter"
## [13] "P_LDPNT: Nationwide percentile score for lead paint indicator (from 0-100)"
## [14] "P_DSLPM: Nationwide percentile score for diesel particulate matter level (from 0-100)"
## [15] "P_CANCR: Nationwide percentile score for cancer risk (from 0-100)"
## [16] "P_RESP: Nationwide percentile score for respiratory hazard index (from 0-100)"
## [17] "P_PTRAF: Nationwide percentile score for proximity to traffic (from 0-100)"
## [18] "P_PWDIS: Nationwide percentile score for major direct dischargers to water (from 0-100)"
## [19] "P_PNPL: Nationwide percentile score for proximity to NPL sites (from 0-100)"
## [20] "P_PRMP: Nationwide percentile score for proximity to RMP facilities (from 0-100)"
## [21] "P_PTSDF: Nationwide percentile score for proximity to TSDF facilities (from 0-100)"
## [22] "P_OZONE: Nationwide percentile score for ozone level (from 0-100)"
## [23] "P_PM25: Nationwide percentile score for PM2.5 level (from 0-100)"
## [24] "T_LDPNT: Map text for lead paint indicator"
## [25] "T_DSLPM: Map text for diesel particulate matter level"
## [26] "T_CANCR: Map text for cancer risk"
## [27] "T_RESP: Map text for respiratory hazard index"
## [28] "T_PTRAF: Map text for proximity to traffic"
## [29] "T_PWDIS: Map text for major direct dischargers to water"
## [30] "T_PNPL: Map text for proximity to NPL sites"
## [31] "T_PRMP: Map text for proximity to RMP facilities"
## [32] "T_PTSDF: Map text for proximity to TSDF facilities"
## [33] "T_OZONE: Map text for ozone level"
## [34] "T_PM25: Map text for PM2.5 level"
## [35] "AREALAND: Land area (in square meters)"
## [36] "AREAWATER: Water area (in square meters)"
ejscreen %>% select(-c(ID, T_LDPNT:T_PM25)) %>%
select(where(~is.numeric(.x) && !is.na(.x))) %>%
as.data.frame() %>%
stargazer(., type = "text", title = "Summary Statistics", digits = 0,
summary.stat = c("mean", "sd", "min", "median", "max"))
##
## Summary Statistics
## ===============================================================
## Statistic Mean St. Dev. Min Median Max
## ---------------------------------------------------------------
## PRE1960PCT 0 0 0 0 1
## DSLPM 0 0 0 0 1
## CANCER 30 3 23 29 34
## RESP 0 0 0 0 0
## PTRAF 248 378 0 67.4 2,184
## PWDIS 0 0 0 0.000 1
## PNPL 0 0 0 0 0
## PRMP 0 0 0 0 1
## PTSDF 1 1 0 0 3
## OZONE 41 0 41 41 42
## PM25 7 0 7 7 8
## P_LDPNT 47 22 11 48 98
## P_DSLPM 34 21 6 25 73
## P_CANCR 43 13 16 40 62
## P_RESP 39 13 13 37 60
## P_PTRAF 34 29 0 29.1 92
## P_PWDIS 30 26 0 41.3 95
## P_PNPL 35 14 16 33 95
## P_PRMP 13 14 2 7 79
## P_PTSDF 34 21 4 32 74
## P_OZONE 38 3 33 39 43
## P_PM25 19 2 13 19 23
## AREALAND 35,769,278 46,646,067 157,587 10,727,976 227,406,673
## AREAWATER 474,082 1,484,460 0 106,800 12,681,090
## ---------------------------------------------------------------
The following charts show the correlations between all combinations of variables. The darker the color, the more highly correlated a pair of variables are. The first correlation matrix shows correlations among the levels of each environmental indicator, and the second shows correlations among the percentiles of each indicator.
correlation <- ejscreen %>%
select(PRE1960PCT:PM25)
num_correlation <- cor(correlation, use = "complete.obs")
num_correlation <- round(num_correlation, digits = 2)
corrplot(num_correlation, type = {"upper"}, method = "shade",
shade.col = NA, tl.col = "black",
diag = F, addCoef.col = "black")
meta %>%
filter(varname %in% c("PRE1960PCT", "DSLPM", "CANCER", "RESP", "PTRAF", "PWDIS", "PNPL", "PRMP", "PTSDF", "OZONE", "PM25")) %>%
mutate(label = paste0(varname, ": ", about)) %>%
select(label) %>%
as.list()
## $label
## [1] "PRE1960PCT: % of housing built before 1960 -- lead paint indicator"
## [2] "DSLPM: Diesel particulate matter level in the air, measured in micrograms per cubic meter"
## [3] "CANCER: Cancer risk due to toxics in the air"
## [4] "RESP: \"Ratio of exposure concentration to health-based reference concentration\""
## [5] "PTRAF: Average number of daily vehicles at major roads divided by distance in meters"
## [6] "PWDIS: Toxicity-weighted stream concentrations divided by distance in kilometers"
## [7] "PNPL: Number of National Priorities List (NPL) sites within 5 km divided by distance in kilometers"
## [8] "PRMP: Number of Risk Management Plan (RMP) facilities within 5 km divided by distance in kilometers"
## [9] "PTSDF: Number of Treatment Storage and Disposal (TSDF) facilities within 5 km divided by distance in kilometers"
## [10] "OZONE: Summer daily average of ozone concentration in the air, in parts per billion"
## [11] "PM25: Yearly average PM2.5 level in the air, measured in micrograms per cubic meter"
correlation2 <- ejscreen %>%
select(P_LDPNT:P_PM25)
num_correlation2 <- cor(correlation2, use = "complete.obs")
num_correlation2 <- round(num_correlation2, digits = 2)
corrplot(num_correlation2, type = {"upper"}, method = "shade",
shade.col = NA, tl.col = "black",
diag = F, addCoef.col = "black")
meta %>%
filter(varname %in% c("P_LDPNT", "P_DSLPM", "P_CANCR", "P_RESP", "P_PTRAF", "P_PWDIS", "P_PNPL", "P_PRMP", "P_PTSDF", "P_OZONE", "P_PM25")) %>%
mutate(label = paste0(varname, ": ", about)) %>%
select(label) %>%
as.list()
## $label
## [1] "P_LDPNT: Nationwide percentile score for lead paint indicator (from 0-100)"
## [2] "P_DSLPM: Nationwide percentile score for diesel particulate matter level (from 0-100)"
## [3] "P_CANCR: Nationwide percentile score for cancer risk (from 0-100)"
## [4] "P_RESP: Nationwide percentile score for respiratory hazard index (from 0-100)"
## [5] "P_PTRAF: Nationwide percentile score for proximity to traffic (from 0-100)"
## [6] "P_PWDIS: Nationwide percentile score for major direct dischargers to water (from 0-100)"
## [7] "P_PNPL: Nationwide percentile score for proximity to NPL sites (from 0-100)"
## [8] "P_PRMP: Nationwide percentile score for proximity to RMP facilities (from 0-100)"
## [9] "P_PTSDF: Nationwide percentile score for proximity to TSDF facilities (from 0-100)"
## [10] "P_OZONE: Nationwide percentile score for ozone level (from 0-100)"
## [11] "P_PM25: Nationwide percentile score for PM2.5 level (from 0-100)"
These scatterplots show the relationship between ozone and PM2.5, broken down by county. The first scatterplot shows the correlation between the levels of ozone and PM2.5, and the second scatterplot shows the correlation between the percentiles.
ejscreen <- ejscreen %>%
mutate(county = str_sub(ID, 3, 5))
ejscreen %>%
ggplot() +
geom_point(aes(x=OZONE, y=PM25, color=county)) +
labs(x="Ozone level",
y="PM2.5 level") +
scale_color_brewer(type = "qual", labels = c("Albemarle", "Fluvanna", "Greene", "Louisa", "Nelson", "Charlottesville"))
ejscreen %>%
ggplot() +
geom_point(aes(x=P_OZONE, y=P_PM25, color=county)) +
labs(x="Ozone percentile",
y="PM2.5 percentile") +
scale_color_brewer(type = "qual", labels = c("Albemarle", "Fluvanna", "Greene", "Louisa", "Nelson", "Charlottesville"))
These scatterplots show the relationship between a block group’s proximity to traffic and air toxics cancer risk, broken down by county. The first one shows the correlation between the levels, and the second one shows the correlation between the percentiles.
ejscreen %>%
ggplot() +
geom_point(aes(x=PTRAF, y=CANCER, color=county)) +
labs(x="Proximity to traffic",
y="Cancer risk") +
scale_color_brewer(type = "qual", labels = c("Albemarle", "Fluvanna", "Greene", "Louisa", "Nelson", "Charlottesville"))
ejscreen %>%
ggplot() +
geom_point(aes(x=P_PTRAF, y=P_CANCR, color=county)) +
labs(x="Traffic proximity percentile",
y="Cancer risk percentile") +
scale_color_brewer(type = "qual", labels = c("Albemarle", "Fluvanna", "Greene", "Louisa", "Nelson", "Charlottesville"))
These scatterplots show the relationship between a block group’s proximity to traffic and its diesel particulate matter level, broken down by county. The first shows the correlation between the levels, and the second shows the correlation between the percentiles.
ejscreen %>%
ggplot() +
geom_point(aes(x=PTRAF, y=DSLPM, color=county)) +
labs(x="Proximity to traffic",
y="Diesel particulate matter level") +
scale_color_brewer(type = "qual", labels = c("Albemarle", "Fluvanna", "Greene", "Louisa", "Nelson", "Charlottesville"))
ejscreen %>%
ggplot() +
geom_point(aes(x=P_PTRAF, y=P_DSLPM, color=county)) +
labs(x="Traffic proximity percentile",
y="Diesel particulate matter percentile") +
scale_color_brewer(type = "qual", labels = c("Albemarle", "Fluvanna", "Greene", "Louisa", "Nelson", "Charlottesville"))
These scatterplots show the relationship between PM2.5 and diesel particulate matter, broken down by county. The first shows the correlation between the levels, and the second shows the correlation between the percentiles.
ejscreen %>%
ggplot() +
geom_point(aes(x=PM25, y=DSLPM, color=county)) +
labs(x="PM2.5 level",
y="Diesel particulate matter level") +
scale_color_brewer(type = "qual", labels = c("Albemarle", "Fluvanna", "Greene", "Louisa", "Nelson", "Charlottesville"))
ejscreen %>%
ggplot() +
geom_point(aes(x=P_PM25, y=P_DSLPM, color=county)) +
labs(x="PM2.5 percentile",
y="Diesel particulate matter percentile") +
scale_color_brewer(type = "qual", labels = c("Albemarle", "Fluvanna", "Greene", "Louisa", "Nelson", "Charlottesville"))
pal <- colorNumeric("Blues", reverse = TRUE, domain = cvilleshapes$PTSDF)
leaflet(cvilleshapes) %>%
addProviderTiles("CartoDB.Positron") %>%
addPolygons(data = cvilleshapes,
fillColor = ~pal(PTSDF),
weight = 1,
opacity = 1,
color = "white",
fillOpacity = 0.6,
highlight = highlightOptions(weight = 2, fillOpacity = 0.8, bringToFront = T),
popup = paste0("FIPS Code: ", cvilleshapes$GEOID, "<br>",
"Proximity to TSDF: ", cvilleshapes$T_PTSDF)) %>%
addLegend("bottomright", pal = pal, values = cvilleshapes$PTSDF,
title = "Proximity to TSDF", opacity = 0.7)
pal <- colorNumeric("Blues", reverse = TRUE, domain = cvilleshapes$PTRAF)
leaflet(cvilleshapes) %>%
addProviderTiles("CartoDB.Positron") %>%
addPolygons(data = cvilleshapes,
fillColor = ~pal(PTRAF),
weight = 1,
opacity = 1,
color = "white",
fillOpacity = 0.6,
highlight = highlightOptions(weight = 2, fillOpacity = 0.8, bringToFront = T),
popup = paste0("FIPS Code: ", cvilleshapes$GEOID, "<br>",
"Proximity to traffic: ", cvilleshapes$T_PTRAF)) %>%
addLegend("bottomright", pal = pal, values = cvilleshapes$PTRAF,
title = "Traffic Proximity", opacity = 0.7)
pal <- colorNumeric("Blues", reverse = TRUE, domain = cvilleshapes$P_PM25)
leaflet(cvilleshapes) %>%
addProviderTiles("CartoDB.Positron") %>%
addPolygons(data = cvilleshapes,
fillColor = ~pal(P_PM25),
weight = 1,
opacity = 1,
color = "white",
fillOpacity = 0.6,
highlight = highlightOptions(weight = 2, fillOpacity = 0.8, bringToFront = T),
popup = paste0("FIPS Code: ", cvilleshapes$GEOID, "<br>",
"PM2.5 Percentile: ", cvilleshapes$P_PM25)) %>%
addLegend("bottomright", pal = pal, values = cvilleshapes$P_PM25,
title = "PM2.5 Percentiles", opacity = 0.7)
pal <- colorNumeric("Blues", reverse = TRUE, domain = cvilleshapes$PM25)
leaflet(cvilleshapes) %>%
addProviderTiles("CartoDB.Positron") %>%
addPolygons(data = cvilleshapes,
fillColor = ~pal(PM25),
weight = 1,
opacity = 1,
color = "white",
fillOpacity = 0.6,
highlight = highlightOptions(weight = 2, fillOpacity = 0.8, bringToFront = T),
popup = paste0("FIPS Code: ", cvilleshapes$GEOID, "<br>",
"PM2.5 Level: ", cvilleshapes$T_PM25)) %>%
addLegend("bottomright", pal = pal, values = cvilleshapes$PM25,
title = "PM2.5 Concentrations", opacity = 0.7)
pal <- colorNumeric("Blues", reverse = TRUE, domain = cvilleshapes$CANCER)
leaflet(cvilleshapes) %>%
addProviderTiles("CartoDB.Positron") %>%
addPolygons(data = cvilleshapes,
fillColor = ~pal(CANCER),
weight = 1,
opacity = 1,
color = "white",
fillOpacity = 0.6,
highlight = highlightOptions(weight = 2, fillOpacity = 0.8, bringToFront = T),
popup = paste0("FIPS Code: ", cvilleshapes$GEOID, "<br>",
"Cancer Risk: ", cvilleshapes$T_CANCR)) %>%
addLegend("bottomright", pal = pal, values = cvilleshapes$CANCER,
title = "Air Toxics Cancer Risk", opacity = 0.7)
pal <- colorNumeric("Blues", reverse = TRUE, domain = cvilleshapes$DSLPM)
leaflet(cvilleshapes) %>%
addProviderTiles("CartoDB.Positron") %>%
addPolygons(data = cvilleshapes,
fillColor = ~pal(DSLPM),
weight = 1,
opacity = 1,
color = "white",
fillOpacity = 0.6,
highlight = highlightOptions(weight = 2, fillOpacity = 0.8, bringToFront = T),
popup = paste0("FIPS Code: ", cvilleshapes$GEOID, "<br>",
"DSLPM: ", cvilleshapes$T_DSLPM)) %>%
addLegend("bottomright", pal = pal, values = cvilleshapes$DSLPM,
title = "Diesel Particulate Matter Level", opacity = 0.7)
pal <- colorNumeric("Blues", reverse = TRUE, domain = cvilleshapes$OZONE)
leaflet(cvilleshapes) %>%
addProviderTiles("CartoDB.Positron") %>%
addPolygons(data = cvilleshapes,
fillColor = ~pal(OZONE),
weight = 1,
opacity = 1,
color = "white",
fillOpacity = 0.6,
highlight = highlightOptions(weight = 2, fillOpacity = 0.8, bringToFront = T),
popup = paste0("FIPS Code: ", cvilleshapes$GEOID, "<br>",
"Ozone Level: ", cvilleshapes$T_OZONE)) %>%
addLegend("bottomright", pal = pal, values = cvilleshapes$OZONE,
title = "Ozone Levels in the Air", opacity = 0.7)
PM2.5, ozone, and NATA indicators (cancer risk, respiratory hazard index, and diesel particulate matter) are measured at the census tract level, and the same value is assigned to each block group within that tract. All other variables were collected at the block group level.